From shishirprakash at live.com Thu Dec 22 15:25:00 2016 Content-Type: multipart/mixed; boundary="===============3449446762197207866==" MIME-Version: 1.0 From: shishir praksh To: users at ovirt.org Subject: [ovirt-users] Allocating only "x" amount of physical memory to hypervisor node Date: Thu, 22 Dec 2016 15:19:06 +0000 Message-ID: --===============3449446762197207866== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable --_000_PN1PR01MB0430420C315967E1A20F3B7CA3920PN1PR01MB0430INDP_ Content-Type: text/plain; charset=3D"iso-8859-1" Content-Transfer-Encoding: quoted-printable The requirement is to use ovirt hypervisor for one application along with v= =3D dsm. I want to only "x" amount of memory should be visible/allocated to hyp= =3D ervisor for hosting the VMs rest I will use for my application hosted on sa= =3D me physical box. pointers will be appreciated !!!! Thanks Shishir- --_000_PN1PR01MB0430420C315967E1A20F3B7CA3920PN1PR01MB0430INDP_ Content-Type: text/html; charset=3D"iso-8859-1" Content-Transfer-Encoding: quoted-printable

The requirement is = =3D to use ovirt hypervisor for one application along with vdsm. I want to only= =3D "x" amount of memory should be visible/allocated to hypervisor for hosting the VMs rest I will use for my application hoste= =3D d on same physical box.
pointers will be appre= =3D ciated !!!!


Thanks

Shishir-

--_000_PN1PR01MB0430420C315967E1A20F3B7CA3920PN1PR01MB0430INDP_-- --===============3449446762197207866== Content-Type: multipart/alternative MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="attachment.bin" LS1fMDAwX1BOMVBSMDFNQjA0MzA0MjBDMzE1OTY3RTFBMjBGM0I3Q0EzOTIwUE4xUFIwMU1CMDQz MElORFBfCkNvbnRlbnQtVHlwZTogdGV4dC9wbGFpbjsgY2hhcnNldD0iaXNvLTg4NTktMSIKQ29u dGVudC1UcmFuc2Zlci1FbmNvZGluZzogcXVvdGVkLXByaW50YWJsZQoKVGhlIHJlcXVpcmVtZW50 IGlzIHRvIHVzZSBvdmlydCBoeXBlcnZpc29yIGZvciBvbmUgYXBwbGljYXRpb24gYWxvbmcgd2l0 aCB2PQpkc20uIEkgd2FudCB0byBvbmx5ICJ4IiBhbW91bnQgb2YgbWVtb3J5IHNob3VsZCBiZSB2 aXNpYmxlL2FsbG9jYXRlZCB0byBoeXA9CmVydmlzb3IgZm9yIGhvc3RpbmcgdGhlIFZNcyByZXN0 IEkgd2lsbCB1c2UgZm9yIG15IGFwcGxpY2F0aW9uIGhvc3RlZCBvbiBzYT0KbWUgcGh5c2ljYWwg Ym94Lgpwb2ludGVycyB3aWxsIGJlIGFwcHJlY2lhdGVkICEhISEKCgpUaGFua3MKClNoaXNoaXIt CgotLV8wMDBfUE4xUFIwMU1CMDQzMDQyMEMzMTU5NjdFMUEyMEYzQjdDQTM5MjBQTjFQUjAxTUIw NDMwSU5EUF8KQ29udGVudC1UeXBlOiB0ZXh0L2h0bWw7IGNoYXJzZXQ9Imlzby04ODU5LTEiCkNv bnRlbnQtVHJhbnNmZXItRW5jb2Rpbmc6IHF1b3RlZC1wcmludGFibGUKCjxodG1sPgo8aGVhZD4K PG1ldGEgaHR0cC1lcXVpdj0zRCJDb250ZW50LVR5cGUiIGNvbnRlbnQ9M0QidGV4dC9odG1sOyBj aGFyc2V0PTNEaXNvLTg4NTktPQoxIj4KPHN0eWxlIHR5cGU9M0QidGV4dC9jc3MiIHN0eWxlPTNE ImRpc3BsYXk6bm9uZTsiPjwhLS0gUCB7bWFyZ2luLXRvcDowO21hcmdpPQpuLWJvdHRvbTowO30g LS0+PC9zdHlsZT4KPC9oZWFkPgo8Ym9keSBkaXI9M0QibHRyIj4KPGRpdiBpZD0zRCJkaXZ0YWdk ZWZhdWx0d3JhcHBlciIgc3R5bGU9M0QiZm9udC1zaXplOjEycHQ7Y29sb3I6IzAwMDAwMDtmb250 PQotZmFtaWx5OkNhbGlicmksQXJpYWwsSGVsdmV0aWNhLHNhbnMtc2VyaWY7IiBkaXI9M0QibHRy Ij4KPHA+PHNwYW4gc3R5bGU9M0QiY29sb3I6IHJnYigyOSwgMzMsIDQxKTsgZm9udC1mYW1pbHk6 ICZxdW90O0hlbHZldGljYSBOZXVlPQomcXVvdDssIEhlbHZldGljYSwgQXJpYWwsIHNhbnMtc2Vy aWY7IGZvbnQtc2l6ZTogMTRweDsiPlRoZSByZXF1aXJlbWVudCBpcyA9CnRvIHVzZSBvdmlydCBo eXBlcnZpc29yIGZvciBvbmUgYXBwbGljYXRpb24gYWxvbmcgd2l0aCB2ZHNtLiBJIHdhbnQgdG8g b25seT0KICZxdW90O3gmcXVvdDsgYW1vdW50IG9mIG1lbW9yeSBzaG91bGQgYmUgdmlzaWJsZS9h bGxvY2F0ZWQKIHRvIGh5cGVydmlzb3IgZm9yIGhvc3RpbmcgdGhlIFZNcyByZXN0IEkgd2lsbCB1 c2UgZm9yIG15IGFwcGxpY2F0aW9uIGhvc3RlPQpkIG9uIHNhbWUgcGh5c2ljYWwgYm94Ljwvc3Bh bj48YnIgc3R5bGU9M0QiY29sb3I6IHJnYigyOSwgMzMsIDQxKTsgZm9udC1mYW09CmlseTogJnF1 b3Q7SGVsdmV0aWNhIE5ldWUmcXVvdDssIEhlbHZldGljYSwgQXJpYWwsIHNhbnMtc2VyaWY7IGZv bnQtc2l6ZTogMT0KNHB4OyI+CjxzcGFuIHN0eWxlPTNEImNvbG9yOiByZ2IoMjksIDMzLCA0MSk7 IGZvbnQtZmFtaWx5OiAmcXVvdDtIZWx2ZXRpY2EgTmV1ZSZxdT0Kb3Q7LCBIZWx2ZXRpY2EsIEFy aWFsLCBzYW5zLXNlcmlmOyBmb250LXNpemU6IDE0cHg7Ij5wb2ludGVycyB3aWxsIGJlIGFwcHJl PQpjaWF0ZWQgISEhITwvc3Bhbj48YnI+CjwvcD4KPHA+PHNwYW4gc3R5bGU9M0QiY29sb3I6IHJn YigyOSwgMzMsIDQxKTsgZm9udC1mYW1pbHk6ICZxdW90O0hlbHZldGljYSBOZXVlPQomcXVvdDss IEhlbHZldGljYSwgQXJpYWwsIHNhbnMtc2VyaWY7IGZvbnQtc2l6ZTogMTRweDsiPjxicj4KPC9z cGFuPjwvcD4KPHA+PHNwYW4gc3R5bGU9M0QiY29sb3I6IHJnYigyOSwgMzMsIDQxKTsgZm9udC1m YW1pbHk6ICZxdW90O0hlbHZldGljYSBOZXVlPQomcXVvdDssIEhlbHZldGljYSwgQXJpYWwsIHNh bnMtc2VyaWY7IGZvbnQtc2l6ZTogMTRweDsiPlRoYW5rczwvc3Bhbj48L3A+CjxwPjxzcGFuIHN0 eWxlPTNEImNvbG9yOiByZ2IoMjksIDMzLCA0MSk7IGZvbnQtZmFtaWx5OiAmcXVvdDtIZWx2ZXRp Y2EgTmV1ZT0KJnF1b3Q7LCBIZWx2ZXRpY2EsIEFyaWFsLCBzYW5zLXNlcmlmOyBmb250LXNpemU6 IDE0cHg7Ij5TaGlzaGlyLTwvc3Bhbj48L3A+CjwvZGl2Pgo8L2JvZHk+CjwvaHRtbD4KCi0tXzAw MF9QTjFQUjAxTUIwNDMwNDIwQzMxNTk2N0UxQTIwRjNCN0NBMzkyMFBOMVBSMDFNQjA0MzBJTkRQ Xy0tCg== --===============3449446762197207866==-- From msivak at redhat.com Fri Dec 23 09:55:59 2016 Content-Type: multipart/mixed; boundary="===============8336346646509747415==" MIME-Version: 1.0 From: Martin Sivak To: users at ovirt.org Subject: Re: [ovirt-users] Allocating only "x" amount of physical memory to hypervisor node Date: Fri, 23 Dec 2016 10:55:58 +0100 Message-ID: In-Reply-To: PN1PR01MB0430420C315967E1A20F3B7CA3920@PN1PR01MB0430.INDPRD01.PROD.OUTLOOK.COM --===============8336346646509747415== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Hi, I believe we do not support this, but there is something similar in the VDSM configuration (/etc/vdsm/vdsm.conf). There are two options: host_mem_reserve and extra_mem_reserve that allow you to reserve memory for the OS and your application. You should be aware of features like fencing though. oVirt can decide to reboot the host and that is probably not what you want when a custom application is running there as well. Is there any reason why the application cannot be run inside a VM? Best regards -- Martin Sivak SLA / oVirt On Thu, Dec 22, 2016 at 4:19 PM, shishir praksh = wrote: > The requirement is to use ovirt hypervisor for one application along with > vdsm. I want to only "x" amount of memory should be visible/allocated to > hypervisor for hosting the VMs rest I will use for my application hosted = on > same physical box. > pointers will be appreciated !!!! > > > Thanks > > Shishir- > > > _______________________________________________ > Users mailing list > Users(a)ovirt.org > http://lists.ovirt.org/mailman/listinfo/users > --===============8336346646509747415==--